Structural Molecular Network for Nontargeted Screening and Prioritization of New Pollutants in Urban Wastewater

High-throughput screening of new
pollutants is critical
for effective
wastewater treatment. However, processing nontargeted high-resolution
mass spectrometry (HRMS) data presents significant challenges. Molecular
networks offer a promising strategy for discovering new pollutants
by constructing networks that link structurally similar compounds.
This study introduces a screening approach that integrates novel structural
molecular networks with HRMS to effectively annotate new pollutants.
This method uses limited known tandem mass spectrometry (MS/MS) compounds
as seeds, enabling the large-scale screening of suspected pollutants
without extensive MS/MS databases. The feasibility of the proposed
method was validated using 319 and 172 pesticides in positive and
negative electrospray ion modes, respectively, achieving annotation
rates exceeding 92.81% and accuracy rates surpassing 83.56%. Notably,
varying the number of seed compounds maintained an annotation accuracy
above 80.16%. Applying this method to the influent and effluent samples
from an urban wastewater treatment plant, 1583 compounds were annotated,
with 232 transformation products being uniquely found in the effluent.
Furthermore, the hazard prioritization of the effluent pollutants
of 74 compounds poses a higher priority. Among these, 8 compounds
are identified as transformation products, and their potential parent
compounds were characterized. This study provides new insights into
the comprehensive monitoring of new pollutants and transformation
products.